PRIVACY-FIRST ANOMALY DETECTION USING RADIO WAVES

The Netherlands is aging rapidly. Not only because the baby boomer generation has been reaching retirement age since 2011, but also because people are living longer, healthier lives. Consequently, many elderly individuals need to continue living independently for a longer period. How can we ensure this happens safely, without placing an additional burden on the healthcare system? Fabian van Prooijen came up with a solution: using radio waves. He reached out to researcher Jeroen Klein Brinke, and together they founded RadioSense: a University of Twente spin-off that uses radio waves to detect significant events, such as fall accidents.

Stairlifts, smart home systems, emergency alarm buttons… There are countless solutions designed to help the elderly live in their own homes for longer. Yet, not all accidents can be prevented, and help is not always immediately available. Fabian witnessed this firsthand with his own grandfather. “My grandfather is 98 years old and still lives independently,” Fabian explains. “He has various aids in his home, including an alarm button. But if he falls, he isn’t always able to press it. On one occasion, he was found behind his front door by the newspaper delivery person, having lain there for several hours. Fortunately, he lived to tell the tale, but in the Netherlands, 100,000 elderly people end up in the emergency room each year following a fall, often suffering severe complications or even death because they weren’t discovered in time. With RadioSense, we want to change that.” 

Ripple effect

The spin-off has developed a plug-and-play device that can be easily installed in a home. Using radio waves and pattern recognition, it can detect whether something unusual is happening within the residence. “You can think of radio waves like the ripple effect in water,” Fabian explains. “When you throw a stone into a pond, ripples form. These ripples change depending on the shape and size of the stone. Radio waves work in a similar way. We can read these waves and use them to identify changes in the pattern. If someone falls, we see a spike in the disruption of the radio signals, followed by a period of complete stillness. At that point, we can send an alert to the relevant care system. We integrate with existing infrastructure and healthcare solutions, allowing care professionals to handle the follow-up.”

From boredom to impact

Although RadioSense’s technology is based on Jeroen’s research, the idea first came to Fabian while he was at a campsite in Italy. “My wife was heavily pregnant during that holiday and took frequent afternoon naps, so I had to find ways to keep myself entertained,” he recalls. “While testing the campsite’s network, I could pinpoint the location of the access point. That got me wondering: could the access point ‘see’ me, too?” Fabian delved into the theory behind the technology and discovered a research paper by Jeroen. “I took a bold step and sent him a message via LinkedIn. Not long after, when my grandfather fell, we decided to use this technology to help the elderly.”

Integration with existing systems

With the help of a TTT voucher, they got to work. That provided the necessary push to establish RadioSense as a UT spin-off. They have since developed a plug-and-play device that uses two smart plugs to create its own Wi-Fi network. By integrating with existing healthcare systems, the invention serves as a valuable addition to current solutions. Furthermore, the system operates without the use of sensitive personal data. Fabian: “We don’t use cameras, and the raw data is processed on the edge (locally rather than being stored, ed.). This makes our solution GDPR-compliant, and the signal cannot be traced back to specific individuals.”

Predictions

Ultimately, Fabian and his team are working toward the full integration of their technology into routers. “For instance, you could detect whether someone who has received a new hip is moving enough to ensure a proper recovery,” says Fabian. “In the long term, we might even be able to predict conditions like Alzheimer’s disease by analyzing the walking patterns a person develops.” The technology also has potential applications beyond healthcare. “For example, by detecting intruders in a port area or securing buildings. It is highly scalable, as it can be deployed within existing infrastructure. But for now, our mission is to help the elderly live at home for longer and lead a dignified life, without compromising their privacy.”